Final Report to the Drinking Water Inspectorate
Report No DWI0725

Feb 1995


Two protocols have been drafted, in BSi format, for the conduct of leaching tests to produce leachates from materials intended for use in contact with drinking water, and for the analysis of these leachates for unspecified organic substances using GCMS. This latter protocol was based on the use of a suite of various types of deuterium-labelled compounds to ensure that the methodology utilised was appropriate to allow the requirements of the protocol to be met.

A within-laboratory performance test of these protocols was undertaken using three materials agreed with the Department. These materials were polyethylene pipe, glass-reinforced polyester (GRP) pipe and bitumen-lined ductile iron pipe. This test demonstrated that, with minimal amendment, the required performance (in terms of limit of detection and relative standard deviations for detected compounds) could be met. The only amendment required was dechlorination of the leachate samples prior to analysis, as it became clear that antioxidants leached from materials could be oxidised by residual free chlorine in the leachates obtained using chlorinated water.

Four laboratories were involved in the subsequent interlaboratory performance testing of the protocols. These were WRc, The Water Quality Centre (Thames Water), KIWA (The Netherlands) and Centre de Recherche et de Controle des Eaux, Paris (CRECEP) (France). This testing showed that comparable results to those from the initial within laboratory testing could be obtained for the internal standards, although difficulties were encountered by two laboratories in detecting the most volatile internal standard. In one case, the chromatographic conditions were such that this internal standard should have been detected, and it therefore appears that the technique used for concentrating the extract prior to GCMS analysis was inappropriate. In the other case, the laboratory concerned did not comply fully with the GCMS protocol. Various criteria, such as the recoveries of the internal standards, the quality of the procedural blanks, the consistency of the ratios of the peak areas of the compounds detected and the internal standards, and the consistency of the mass spectra obtained were used to evaluate the performance of the protocols.

Provided they are strictly adhered to, the protocols are satisfactory, in that they allowed organic compounds added to leachates from materials at known levels to be reliably determined at concentrations in the range 1-10 µg/l. The limit of detection was about 0.75 µg/l. However, for compounds leached from the materials being tested the performance was more difficult to assess and there were discrepancies between laboratories, both in terms of compounds detected and in quantifying these compounds. Some of discrepancies may have been due to an increased variation arising from the leaching process itself, and some problems were caused by difficulties in assessing whether compounds detected in both leachates and blanks were classified as originating from the materials tested.

Some recommendations are given regarding possible improvements to the protocols.

Copies of this report may be available as an Acrobat pdf download under the 'Pre 2000 Reports' heading on the DWI website.